Tír Na HiFi

lekt.exe1644 v2.6

Now it's working.
Surprisingly good! Analog and tube amp signature. Like it so much.

Thank you.

John Dot wrote:lekt.exe1644 v2.6

Now it's working.
Surprisingly good! Analog and tube amp signature. Like it so much.

Thank you.

I think v1.8_v2 have better bass (multi-layer) than v2.6, so sound weigth and depth seem more correct. That play by reading from file is not bad method as people think, in this case i see better, v1.8_v2 used interesting read file method. maybe let pagefile.sys work in windows coz it well support for reading file. my Win8 use pagefile.sys
I will try more working with RAM versions.

lekt wrote:

John Dot wrote:lekt.exe1644 v2.6

Now it's working.
Surprisingly good! Analog and tube amp signature. Like it so much.

Thank you.

I think v1.8_v2 have better bass (multi-layer) than v2.6, so sound weigth and depth seem more correct. That play by reading from file is not bad method as people think, in this case i see better, v1.8_v2 used interesting read file method. maybe let pagefile.sys work in windows coz it well support for reading file. my Win8 use pagefile.sys
I will try more working with RAM versions.

don't like v2.6

1.8_v2
-finally plays all files. no bugs!
-better better. emotion is kicking in!! need more more vibration/micro-details in vocal.
-loving the tune and vibration.
-stage is great.
-very mellow+good vibration on easily-hard-trebles. just good.
-sounds great on my new installed 7. also great on R2.
-earaches is close to nothing on headphones.
-deeper fuller than 1.8.1 and 1.8.0

(just installed VS2013 on 7. this ".NET Runtime Optimization Service" takes forvever to go away. just annoying)

jesuscheung wrote:...(just installed VS2013 on 7. this ".NET Runtime Optimization Service" takes forvever to go away. just annoying)

heared XA's win7, interesting. win7 have low latency, easier setup. but most important versions audiodg.exe, AudioEndpointBuilder.dll, audiosrv.dll,... for win8 are good programing, i think Microsort take more care about these. in feature seem to be win8.1 is too, background audio services must better. i'll try its wasapi.
currently wasapi have many, many things not good for render loop (error checking, return values, not use static variables,...), bad characteristic for realtime asynchronous transfer data. im probably wrong.

lekt wrote:

jesuscheung wrote:...(just installed VS2013 on 7. this ".NET Runtime Optimization Service" takes forvever to go away. just annoying)

heared XA's win7, interesting. win7 have low latency, easier setup. but most important versions audiodg.exe, AudioEndpointBuilder.dll, audiosrv.dll,... for win8 are good programing, i think Microsort take more care about these. in feature seem to be win8.1 is too, background audio services must better. i'll try its wasapi.
currently wasapi have many, many things not good for render loop (error checking, return values, not use static variables,...), bad characteristic for realtime asynchronous transfer data. im probably wrong.

if lekt isn't wasapi what does it use ?

how do you know audiodg.exe, AudioEndpointBuilder.dll, audiosrv.dll,... for win8 are good programing ? Do you have a link ?

currently wasapi have many, many things not good for render loop (error checking, return values, not use static variables,...), that's very similar to what I put in the MQn thread, great minds think alike, eh.

sbgk wrote:if lekt isn't wasapi what does it use ?

how do you know audiodg.exe, AudioEndpointBuilder.dll, audiosrv.dll,... for win8 are good programing ? Do you have a link ?

currently wasapi have many, many things not good for render loop (error checking, return values, not use static variables,...), that's very similar to what I put in the MQn thread, great minds think alike, eh.

exclusive mode, but on each step of loop client must transfer &nFramesInBuffer, &pData,... What is this, sbgk. there's like as shared mode. all time of playing use only 2 buffers: 1 for DAC, 1 for clients loading data, and then swap roles. but pData pointers it variable, not fixed. why? i think about how can i don't use GetBuffer(...). 3 WASAPIs function in render loop contains many, many inside works that i don't need (checking error, return many hr values,...). thanks to Microdoft but sorry Microsoft, i'm waiting for other WASSPIs version. computer player must be too much better, it's really, sbgk.

lekt wrote:

sbgk wrote:if lekt isn't wasapi what does it use ?

how do you know audiodg.exe, AudioEndpointBuilder.dll, audiosrv.dll,... for win8 are good programing ? Do you have a link ?

currently wasapi have many, many things not good for render loop (error checking, return values, not use static variables,...), that's very similar to what I put in the MQn thread, great minds think alike, eh.

exclusive mode, but on each step of loop client must transfer &nFramesInBuffer, &pData,... What is this, sbgk. there's like as shared mode. all time of playing use only 2 buffers: 1 for DAC, 1 for clients loading data, and then swap roles. but pData pointers it variable, not fixed. why? i think about how can i don't use GetBuffer(...). 3 WASAPIs function in render loop contains many, many inside works that i don't need (checking error, return many hr values,...). thanks to Microdoft but sorry Microsoft, i'm waiting for other WASSPIs version. computer player must be too much better, it's really, sbgk.

why can't just be like the old days, read hard disk bit by bit bypassing the cache, send data to memory to DAC bypassing the OS control. everywhere i see there is a buffer from disk, CPU, player, dac etc. remember when i was kid playing DOS there was virtually no error checking, error was BSOD.

lekt wrote:

sbgk wrote:if lekt isn't wasapi what does it use ?

how do you know audiodg.exe, AudioEndpointBuilder.dll, audiosrv.dll,... for win8 are good programing ? Do you have a link ?

currently wasapi have many, many things not good for render loop (error checking, return values, not use static variables,...), that's very similar to what I put in the MQn thread, great minds think alike, eh.

exclusive mode, but on each step of loop client must transfer &nFramesInBuffer, &pData,... What is this, sbgk. there's like as shared mode. all time of playing use only 2 buffers: 1 for DAC, 1 for clients loading data, and then swap roles. but pData pointers it variable, not fixed. why? i think about how can i don't use GetBuffer(...). 3 WASAPIs function in render loop contains many, many inside works that i don't need (checking error, return many hr values,...). thanks to Microdoft but sorry Microsoft, i'm waiting for other WASSPIs version. computer player must be too much better, it's really, sbgk.

you answer my questions and I'll answer yours, that's fair, don't you think.

jesuscheung wrote:why can't just be like the old days, read hard disk bit by bit bypassing the cache, send data to memory to DAC bypassing the OS control. everywhere i see there is a buffer from disk, CPU, player, dac etc. remember when i was kid playing DOS there was virtually no error checking, error was BSOD.

SSD difference HDD, i think SDD don't use chunk size 4096, maybe buffer, cache... is too, not sure. SSD can perfectly read from random address, disk deffractment seem no affect.
lekt v1.7, v1.8x have benefit with HDD, but with SSD not sure. i need to learn SSD technologue, maybe very interesting for reading file, JC.

lekt wrote:

jesuscheung wrote:why can't just be like the old days, read hard disk bit by bit bypassing the cache, send data to memory to DAC bypassing the OS control. everywhere i see there is a buffer from disk, CPU, player, dac etc. remember when i was kid playing DOS there was virtually no error checking, error was BSOD.

SSD difference HDD, i think SDD don't use chunk size 4096, maybe buffer, cache... is too, not sure. SSD can perfectly read from random address, disk deffractment seem no affect.
lekt v1.7, v1.8x have benefit with HDD, but with SSD not sure. i need to learn SSD technologue, maybe very interesting for reading file, JC.

i have HDD.

do you know there is a freeware called RAMDisk 4.3.0 RC 1? could this be faster than SSD?